Buy-one-get-one-free deals attract more attention than percentage deals

UvA-DARE is a service provided by the library of the University of Amsterdam (http

://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

Gordon-Hecker, T.; Pittarello, A.; Shalvi, S.; Roskes, M.

DOI

10.1016/j.jbusres.2019.02.070

Publication date

2020

Document Version

Final published version

Published in

Journal of Business Research

License

Article 25fa Dutch Copyright Act (https://www.openaccess.nl/en/in-the-netherlands/you-share-

we-take-care)

Link to publication

Citation for published version (APA):

Gordon-Hecker, T., Pittarello, A., Shalvi, S., & Roskes, M. (2020). Buy-one-get-one-free deals

attract more attention than percentage deals.

Journal of Business Research

111

, 128-134.

https://doi.org/10.1016/j.jbusres.2019.02.070

General rights

It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s)

and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open

content license (like Creative Commons).

Disclaimer/Complaints regulations

If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please

let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material

inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter

to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You

will be contacted as soon as possible.

Download date:13 Sep 2024

Contents lists available at ScienceDirect

Journal of Business Research

journal homepage: www.elsevier.com/locate/jbusres

Buy-one-get-one-free deals attract more attention than percentage deals

☆

Tom Gordon-Hecker

⁎

, Andrea Pittarello

, Shaul Shalvi

, Marieke Roskes

Ben-Gurion University of the Negev, Israel

Brooklyn College, City University of New York, United States of America

University of Amsterdam, Netherlands

VU University Amsterdam, Netherlands

ARTICLE INFO

Keywords:

Attention

Promotion deals

Preferences

Online commerce

ABSTRACT

Promotion deals and price reductions are common strategies retailers use to attract consumers. We investigate

which of two common types of deals better captures consumers' attention. By tracing eye movements, we ex-

amine participants' attention allocation when deciding between “buy-one-get-one free” (BOGO) deals versus

deals that oﬀer an equivalent price reduction. Results show that people prefer BOGO deals, and they tend to

choose them over price reductions even when the deals are equal in terms of net value. The preference is

ampliﬁed when the discount is relatively high: In these cases, BOGO deals attract more attention than percentage

deals. Overall, our ﬁndings can help retailers develop promotional strategies to capture potential consumers'

attention in online commerce. At the same time, our results warn consumers to better evaluate their options and

not be lured by the ﬁrst BOGO deal that captures their attention, as it might not be the best deal available.

1. Introduction

Online commerce is an ever-increasing industry with an annual

growth rate of > 16%. In the second quarter of 2017—in the United

States alone—the total revenue of online commerce totaled $105.1

billion, which accounted for 8.2% of the total sales in that period (U.S.

Department of Commerce, 2017). As online commerce grows, so does

the competition: Well-known companies (e.g., Amazon, eBay, Groupon)

as well as small businesses are continuously striving to devise novel

promotional tools to increase their market share and boost ﬁnancial

returns. Among the diﬀerent strategies that companies have at their

disposal to lure customers into higher and more frequent purchases,

promotion deals are perhaps the most widely used (Bogomolova, Dunn,

Trinh, Taylor, & Volpe, 2015). These strategies tend to be well received

by customers (Blattberg & Neslin, 1989), and because of their potential,

it is important for retailers to better understand how to tailor them to

maximize their impact of sales. We examine how two diﬀerent pro-

motion deals, namely, buy-one-get-one-free and discount deals, attract

consumers' attention when presented side by side on a computer screen

– a setting that closely resembles the online purchasing experience.

Furthermore, we look at whether the preferences for, and attention

allocation to, such promotion deals diﬀer across discount levels.

2. Theoretical background

2.1. Promotion deals

One of the most prominent ways retailers can attract consumers to

purchase their products is to oﬀer discount deals. Indeed, promotional

deals increase buying intentions by enhancing the value of a certain

product (Grewal, Krishnan, Baker, & Borin, 1998; Lattin & Bucklin,

1989). Deals can generally be classiﬁed into two types: Some add a free

product upon the purchase of some quantity of either related or un-

related products. Such deals are called “buy-one-get-one-free” (BOGO)

deals, “free gifts”, or “bundle oﬀers”, and refer to promotions such as

“buy a necklace and get earrings for free” (Raghubir, 2004; for further

reading on the psychology of “free gifts” see Shampanier, Mazar, &

Ariely, 2007; Mazar, Shampanier, & Ariely, 2016).

Other deals use a percentage discount of the original price, such as

“50% oﬀ” (Sinha & Smith, 2000). From a rational standpoint, the

market value of a product should not be aﬀected by the type of deal. Put

diﬀerently, a rational consumer should be indiﬀerent between a BOGO

deal and another deal oﬀering a 50% discount upon the purchase of two

items. However, research on framing eﬀects and prospect theory

(Kahneman & Tversky, 1979; Tversky & Kahneman, 1974; but see Gal &

Rucker, 2018) proposes that people consider the two oﬀers diﬀerently:

https://doi.org/10.1016/j.jbusres.2019.02.070

Received 15 July 2018; Received in revised form 16 February 2019; Accepted 28 February 2019

☆

This work was supported by the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation program, grant agreement

ERC-StG-637915 and the Netherlands Organization for Scientiﬁc Research, Veni grant 451-15-030.

⁎

Corresponding author at: Department of Psychology, Ben Gurion, University of the Negev, Beer-Sheva 84105, Israel.

E-mail address: [email protected] (T. Gordon-Hecker).

Journal of Business Research 111 (2020) 128–134

Available online 09 March 2019

from the consumer's reference point, BOGO deals are perceived in terms

of additional gains, whereas discount deals are perceived in the form of

reduced losses (Diamond & Sanyal, 1990).

Several studies have examined the eﬀect of these two types of deals

on consumers' preferences (see Krishna, Briesch, Lehmann, & Yuan,

2002, for a meta-analysis) and evaluations of the products oﬀered. For

instance, Raghubir (2004) suggested the value-discounting hypothesis,

according to which price bundles decrease the perceived value of the

product oﬀered as a gift and result in lower willingness to pay for the

free gift as a stand-alone product. In line with these ﬁndings, and ac-

cording to prospect theory's (Kahneman & Tversky, 1979) predictions,

overall, people tend to prefer percentage deals to BOGO deals (Diamond

& Sanyal, 1990). However, this preference reverses when a percentage

deal is a multi-buy deal as well, i.e., when it requires the purchase of

some quantity of the product (e.g. “buy 2 and get 50% oﬀ”). In this

situation, BOGO deals are preferred over percentage deals, arguably

because the percentage deals include an aversive purchasing compul-

sion (Sinha & Smith, 2000). Furthermore, children tend to prefer BOGO

deals even when they are objectively inferior to percentage deals

(Boland, Connell, & Erickson, 2012). Lastly, when price reduction is

large, people prefer BOGO deals to package enlargement that results in

equal savings (Hardesty & Bearden, 2003).

Clearly, diﬀerent deals aﬀect people's preferences and behavior

diﬀerently. However, little is known about the cognitive processes as-

sociated with these preferences. Those include for instance the number

of gazes people devote to each type of deal, which may be considered a

proxy of their preference. Understanding these processes can be useful

for retailers when designing promotion deals and trying to stand out

among the vast range of products from which consumers need to

choose. Here, we use eye-tracking methodologies to examine whether

BOGO deals capture more attention than percentage deals, whether this

depends on discount level, and whether this relates to the choices that

people make.

2.2. Consumer behavior and attention allocation

In recent years, eye-tracking measures have been extensively used

as a measurement of implicit attention and motivational processes

(Fiedler, Glöckner, Nicklisch, & Dickert, 2013; Orquin & Mueller-Loose,

2013). The general ﬁnding is that people tend to ﬁxate longer on the

option they end up choosing (Shimojo, Simion, Shimojo, & Scheier,

2003), and that their attention shifts toward the more favorable or

preferred option (Balcetis & Dunning, 2006; Gable & Harmon-Jones,

2010). For example, studies have found that dieters look longer at foods

(Papies, Stroebe, & Aarts, 2008), smokers at cigarettes (Mogg, Bradley,

Field, & De Houwer, 2003), and heavy drinkers at alcohol (Townshend

& Duka, 2001). This attention shift inﬂuences subsequent behavior. For

instance, when given an option to proﬁt from misreporting the outcome

of a die roll, people who ﬁxated longer on irrelevant yet tempting in-

formation (i.e., an outcome diﬀerent than the one they were instructed

to report) were more likely to use this information to increase their

payoﬀ (Pittarello, Leib, Gordon-Hecker, & Shalvi, 2015).

In the marketing domain, eye-tracking techniques have been used to

provide valuable insights into consumer behavior. For example, the

location of products on supermarket shelves inﬂuences consumers' be-

havior through visual attention (Chandon, Hutchinson, Bradlow, &

Young, 2009). Similarly, click-through rates on internet banners have

decreased because they fail to attract people's attention (Drèze &

Hussherr, 2003). Conversely, people look longer and more often at

salient and self-relevant information (Bee, Prendinger, Nakasone,

André, & Ishizuka, 2006; Glöckner, Fiedler, Hochman, Ayal, & Hilbig,

2012).

We conducted two experiments to examine how people weigh dif-

ferent types of promotion deals appearing on a computer screen – a

setting resembling online shopping. We asked participants to choose

between two deals—one presented as a BOGO deal and one as a

percentage deal. To make the deals as equal as possible, both types of

deals were presented as multi-buy deals. In the experimental trials, the

net value - as well as the total number of products being purchased - of

the two deals was identical, whereas in the ﬁller trials, one deal was

objectively better than the other. For example, in an experimental trial,

participants had to choose between a “buy 2 get 1 free” deal, and a “buy

3 get 33% oﬀ” deal. Conversely, in a ﬁller trial participants had to

choose between a “buy 3 get 2 free” deal, and a “buy 5 get 20% oﬀ”

deal. We tracked participants' eye movements during the decision phase

to examine how their attention was allocated between the two alter-

natives.

The ﬁrst goal of the current experiments was to replicate previous

ﬁndings showing that when choosing between a BOGO deal and a

multi-buy percentage deal (i.e., a percentage deal that requires pur-

chasing some quantity of items), people prefer BOGO over percentage

deals (Sinha & Smith, 2000). Our second goal was to examine how these

deals shape participants' attention, here measured via eye movements.

Because attractive information draws attention (Meißner, Musalem, &

Huber, 2016), we predicted that people would look more at BOGO deals

than at percentage deals. Additionally, we explored how diﬀerent levels

of discounts moderate the predicted eﬀect.

All data are available on the Open Science Framework (https://osf.

io/5qk8y/?view_only=ec31136919474e7ebbfccb5703b37d73). No

participants were excluded from any of the analyses, and all in-

dependent variables and manipulations are reported in the Method

section and the SOM.

3. Experiment 1

3.1. Method

3.1.1. Participants

Thirty undergraduate students from an Israeli university (20 fe-

males, M

age

= 25.13, SD

age

= 1.74) participated in the experiment in

exchange for a show-up fee of 25 ILS (~$7).

3.1.2. Measures and procedure

Participants were seated in a private cubicle, 60 cm from a 24-in

computer monitor (1280 × 1024 maximal resolution). Eye movements

were recorded using a Tobii T120 Eye Tracker (sampling rate = 120 Hz;

accuracy = 0.45°), with a standard 9-point eye-tracking calibration.

Participants were presented with 96 trials in which they had to

choose between two brands of the same product (e.g., cookies,

mayonnaise) each associated with a diﬀerent deal. Participants were

instructed not to choose based on their favorite brand but only ac-

cording to the deal they thought was ﬁnancially better. Speciﬁcally, the

verbatim instructions were “Choose the product that represents the deal

you think is better”. To increase engagement in the task, participants

were told that at the end of the experiment one trial would by randomly

chosen, and if they correctly picked the better deal, they would receive

additional 5 ILS (~$1.5).

In each trial, one deal was a percentage discount (“buy 2, get 50%

oﬀ”) and the other was a BOGO discount (“buy 1 get 1 for free”, see

Fig. 1; for the complete list of trials, see the Appendix A). Because a

recent review of eye-tracking studies (Orquin & Mueller-Loose, 2013)

identiﬁed stimulus characteristics that inﬂuence gaze behavior (i.e.,

saliency, size, visual clutter, and location), in the current experiment,

we sought to control for those characteristics to rule out possible arti-

facts. Hence, the presented deals were similar in terms of word count

and reading time in order to limit any eﬀect of presentation format on

eye movements. Speciﬁcally, both deals consisted of three words and

two numeric values, were located the same distance from the center of

the screen, and their order of presentation was counterbalanced. Fi-

nally, BOGO deals consisted of 16 characters, and percentage deals

consisted of 18 characters. Although sentence length (in textual terms)

inﬂuences ﬁxation duration (see Pieters & Wedel, 2004; Reichle,

T. Gordon-Hecker, et al.

Journal of Business Research 111 (2020) 128–134

129

Rayner, & Pollatsek, 2003), the minor diﬀerence in character quantity

between the two types of deals provides a conservative test for the idea

that people ﬁxate more on the BOGO deals (which consisted of fewer

characters) than on the percentage deals (which consisted of more

characters). If sentence length would drive our results, rather than type

of deal, we should ﬁnd more ﬁxations on percentage deals than on

BOGO deals.

In each trial, a ﬁxation cross appeared on the screen for 500 milli-

seconds. Next, the ﬁxation cross disappeared, and the two products and

their deals were presented respectively on the right and on the left side

of the ﬁxation cross for 6 s. Then the products disappeared, and parti-

cipants were asked to press the key (using the keyboard) corresponding

to their chosen deal. Once participants responded, the next trial began.

Of the total 96 trials, 32 were experimental trials in which the two

deals were equal in terms of value (e.g., “buy 5 and get 40% oﬀ” vs.

“buy 3 and get 2 for free”). To make sure participants would keep trying

to identify the ﬁnancially “best deal,” the other 64 trials were ﬁller

trials in which one deal was better than the other. Within subjects, we

manipulated the magnitude of the deals, so that participants were

presented with deals that yielded a discount of 20%, 25%, 33%, or 40%.

We counterbalanced the order of presentation of the deals (left to right)

and the matching of each deal to each product.

After completing the task, participants completed the approach-

avoidance temperament questionnaire (Elliot & Thrash, 2010), de-

signed to assess their approach (α = 0.795) and avoidance (α = 0.819)

motivation (see all items in the supplementary materials). The partici-

pants indicated on a 7-point Likert scale the extent to which they agreed

with 12 statements. For example, “By nature, I am a very nervous

person” (avoidance) and “Thinking about the things I want really en-

ergizes me” (approach). Additionally, participants completed the 10-

item Rational-Experiential inventory (REI-10; Epstein, Pacini, Denes-

Raj, & Heier, 1996), which assesses people's thinking styles (intuitive or

rational thinking) by means of two sub-scales (Faith in Intuition,

α = 0.784 and Need for Cognition, α = 0.743). Participants rated on a 5-

point scale the extent to which 10 items (ﬁve for each subscale) were

characteristic of them (1 = Very unlike me and 5 = Very like me). Be-

cause these scales were exploratory and outside the main goal of our

current research, we report their results in the supplementary materials.

3.1.3. Validity threats

In the current experiment, we took measures to reduce potential

validity threats that are common in eye-tracking studies (Orquin &

Holmqvist, 2018): To avoid making inappropriate comparisons, we

used a large set of stimuli (i.e., 32 diﬀerent stimuli) that are very similar

to each other in terms of length, size, and character count, and diﬀer

only on the relevant features (i.e., the type of deal). Using a large set of

stimuli also reduces to a minimum the threat of under sampling and the

eﬀects of random stimulus features. To avoid drawing unwarranted

conclusions from multiple analyses, we only make a few, speciﬁc, and

theory-driven ex-ante comparisons. To increase data quality and cap-

ture rate, we used rather large areas of interest (358 × 406 pixels), with

a relatively low width-to-length ratio (1.13), which was shown to im-

prove capture rate (Orquin & Holmqvist, 2018). Finally, to avoid

threats of peripheral processing, which may result in processing the

stimulus without ﬁxating on it, the stimuli appeared relatively distant

from each other, with a distance of 218 pixels (5.77 cm) between the

closest edges.

3.2. Results

3.2.1. Filler trials

In the ﬁller trials, where one deal was objectively better than the

other, participants selected the ﬁnancially better deal in 78.4% of the

trials (1505 of 1920), which was greater than chance, χ

= 336.51,

p < .001. Participants more often selected the ﬁnancially worse deal

when the BOGO deal was worse than the percentage deal (32.40%

chose BOGO, 311 of 960) compared with situations in which the BOGO

deal was better than the percentage deal (10.83% chose percentage

deal, 104 of 960). A generalized linear mixed model logistic regression

controlling for the random eﬀect of participants, with type of the better

deal predicting participants' accuracy revealed that this diﬀerence was

highly signiﬁcant, F(1, 1918) = 128.37, p < .001, b = 1.46, 95%

CI = [1.206, 1.711].

3.2.2. Percentage vs. BOGO deals

In the experimental trials, where the two types of deals were of

equal net value, we replicated previous ﬁndings (Sinha & Smith, 2000),

as participants selected BOGO deals (74.89%, 719 of 960) over per-

centage deals (25.10%, 241 of 960), which was signiﬁcantly diﬀerent

from a non-preferential 50%–50% distribution, χ

= 126.86, p < .001.

3.2.3. Discount level

To enable us to take into account the eﬀects of discount magnitude,

we split the discount levels into low discounts (20% and 25%) and high

discounts (33% and 40%). We ran a generalized linear mixed model

logistic regression controlling for the random eﬀect of participants,

with the discount level predicting participants' selected deal. The ana-

lysis revealed that discount level inﬂuenced selections, with partici-

pants being more likely to select BOGO deals when the discount was

high (77.71%, 373 of 480) than when it was low (72.08%, 346 of 480),

F(1, 958) = 4.72, p = .030.

3.2.4. Gaze behavior

We ran a generalized linear mixed model controlling for the random

eﬀect of participants, with discount level (high vs. low), deal type

(BOGO vs. percentage), and their interaction predicting ﬁxation count

on each of the two products in the experimental trials. The analysis

revealed a deal type × discount level interaction, F(1, 1792) = 24.33,

p < .001 (see Fig. 2). When the discount level was high, participants

exhibited more ﬁxations on BOGO deals (M = 9.73, SD = 3.35) than on

percentage deals (M = 8.85, SD = 3.47), F(1, 1792) = 15.85,

p < .001, b = 0.87, 95% CI = [0.441, 1.299]. However, when the

discount level was low, participants exhibited fewer ﬁxations on BOGO

deals (M = 9.04, SD = 3.20) than on percentage deals (M = 9.69,

SD = 3.58), F(1, 1792) = 8.94, p = .003, b = −0.65, 95%

CI = [−0.223, −1.073]. The main eﬀects for deal type and discount

level were not signiﬁcant (F's < 1).

Fig. 1. Example of an experimental trial in Experiment 1.

T. Gordon-Hecker, et al.

Journal of Business Research 111 (2020) 128–134

130

3.3. Discussion

Experiment 1 revealed that people ﬁxate more often on percentage

deals than on BOGO deals when the discount is low. When discount

level is high, people ﬁxate more often on BOGO deals than on per-

centage deals. Additionally, BOGO deals tend to be chosen over per-

centage deals, especially when the discount level is high. Taken to-

gether, it seems that when the net value of the discount increases,

BOGO deals become more attractive than percentage deals.

However, Experiment 1 has some limitations challenging the in-

terpretation of the results. First, the instructions might have been

somewhat confusing for participants, since we asked participants to

choose the ﬁnancially best deal, when in fact, in the experimental trials

an objectively better deal was not available. Further, participants'

choices had no consequences, potentially limiting their motivation to

exert eﬀort and carefully compare the two deals in the experimental

trials, which may pose a threat to the external validity of our ﬁndings.

Finally, participants were always exposed to the two deals for a ﬁxed

time of 6 s. Fixing exposure time has advantages, because it ensures that

participants can take enough time to process both stimuli. However, it

can also complicate the interpretation of the data, because participants

may adjust their gazes to ﬁt the time window which may attenuate gaze

diﬀerences between options (Orquin & Holmqvist, 2018). Finally, when

people shop online they can usually look at deals as long as they want;

ﬁxed exposure times therefore pose a threat to external validity.

We devised Experiment 2 to address these limitations, and with the

goal of replicating the results obtained in Experiment 1 using a larger

sample. We made the following modiﬁcations: (1) asking participants to

choose the deal they prefer, instead of choosing the ‘best deal’, (2) in-

centivizing participants' choices, (3) having an unlimited exposure time

while the deals were presented, (4) providing choices between pairs of

identical products with diﬀerent deals, to avoid inﬂuence of preference

for speciﬁc products, and to increase similarity to real purchasing si-

tuations where people look for the best deal for a speciﬁc product, and

(5) controlling for participants' numeracy skills (Weller et al., 2013)

which might aﬀect their ability to properly compare the diﬀerent deals

(Tan & Bogomolova, 2016). Additionally, those modiﬁcations allowed

us to examine purchasing behavior that more closely resembles online

shopping. When consumers shop online, they often look for a speciﬁc

product and compare prices oﬀered by multiple retailers, rather than

looking for diﬀerent products from the same retailer. Thus the task used

in Experiment 2 mimics this online shopping experience.

4. Experiment 2

4.1. Method

4.1.1. Participants

Fifty undergraduate students from an Israeli university (25 females,

age

= 24.72, SD

age

= 2.35) participated in the experiment in ex-

change for a show-up fee of 25 ILS (~$7).

4.1.2. Measures and procedure

Participants were seated in a private cubicle, 60 cm from a 24-in

computer monitor (1280 × 1024 maximal resolution). Eye movements

were recorded using a Tobii T120 Eye Tracker (sampling rate = 120 Hz;

accuracy = 0.45°), with a standard 9-point eye-tracking calibration.

The task was similar to that of Experiment 1, with several mod-

iﬁcations. Participants were presented with 96 trials in which they had

to choose between two proposed deals to purchase one of four types of

chocolate bars. Each trial presented two identical chocolate bars, each

representing a diﬀerent deal. Speciﬁcally, the participants were asked

to choose the deal that they preferred, using the keyboard. In each trial,

one deal was a multi-buy percentage discount (“buy 2, get 50% oﬀ”)

and the other was a BOGO discount (“buy 1 get 1 for free”, see Fig. 3;

see the Appendix A for the complete list of trials). To incentivize par-

ticipants, after the completion of the experiment, one trial was ran-

domly selected and participants could purchase the product displayed

on the corresponding trial according to the deal they preferred. As in

Experiment 1, the presented deals were similar in terms of word count,

Fig. 2. Total number of ﬁxations on each of the deal types (BOGO vs. percentage) in the two discount-level conditions in Experiment 1. Error bars represent 95% CI.

Fig. 3. Example of an experimental trial in

Experiment 2.

T. Gordon-Hecker, et al.

Journal of Business Research 111 (2020) 128–134

131

character count and reading time in order to limit any eﬀects of pre-

sentation format on eye movements.

In each trial, a ﬁxation cross appeared on the screen for 500 milli-

seconds. Next, the ﬁxation cross disappeared, and the two deals were

presented on the right and on the left side of the ﬁxation cross until

participants gave a response by pressing the key corresponding to their

preferred deal. Once participants made their choice, the next trial

began.

Of the total 96 trials, 32 were experimental trials in which the two

deals were equal in terms of value (e.g., “buy 5 and get 40% oﬀ” vs.

“buy 3 and get 2 for free”). To keep participants engaged throughout

the task, the other 64 trials were ﬁller trials in which one deal was

better than the other. Within subjects, we manipulated the magnitude

of the deals, so that participants were presented with deals that yielded

a discount of 20%, 25%, 33%, or 40%. We counterbalanced the order of

presentation of the deals (left to right) and the matching of each deal to

each product.

After completing the task, participants completed the Abbreviated

Numeracy Scale (Weller et al., 2013) to assess their numeral abilities.

The scale includes eight items such as: “If the chance of getting a dis-

ease is 10%, how many people would be expected to get the disease out

of 1000?”. Additionally, to control for liking eﬀects, participants in-

dicated what was the maximum price they were willing to pay for each

of the products presented in the experiment.

4.1.3. Validity threats

We employed the same measures as in Experiment 1 to reduce po-

tential validity threats, with two diﬀerences. First, in Experiment 2 the

size of the areas of interest was 298 × 566 pixels, yielding a width-to-

length ratio of 1.90, which is slightly higher than in Experiment 1, but

still allows for a good capture rate (Orquin & Holmqvist, 2018). Second,

the distance between the closest edges of the stimuli was 286 pixels

(7.57 cm), which is larger than the distance in Experiment 1, reducing

the threat of peripheral processing even further. All other measures we

used in Experiment 1 were also implemented in Experiment 2.

4.2. Results

4.2.1. Filler trials

As in Experiment 1, in the ﬁller trials, where one deal was objec-

tively better than the other, participants preferred the better deal in

73.94% of the trials (2366 of 3200), which was signiﬁcantly greater

than chance, χ

= 389.01, p < .001. Participants more often chose the

ﬁnancially worse deal when the percentage deal was better than the

BOGO deal (34.88% chose BOGO, 558 of 1600) compared with trials in

which the BOGO deal was better than the percentage deal (17.25%

chose percentage deal, 276 of 1600). A generalized linear mixed model

logistic regression controlling for the random eﬀect of participants,

with type of the better deal predicting participants' accuracy revealed

that this diﬀerence was highly signiﬁcant, F(1, 3198) = 134.25,

p < .001, b = 1.02, 95% CI = [0.849, 1.195]. As a robustness check,

we entered into the model participants' numeracy score and willingness

to pay for the diﬀerent products in the experiment. This did not aﬀect

the pattern of the results. The analysis further revealed that numeracy

was positively associated with choosing the ﬁnancially best deals, F(1,

3193) = 6.02, p = .014, b = 0.16, 95% CI = [0.032, 0.283]. Finally,

the willingness to pay for the diﬀerent products had no eﬀect on the

number of times people chose the ﬁnancially worse deal (F's < 1.13,

p's > .343).

4.2.2. Percentage vs. BOGO deals

In the experimental trials, where the two types of deals were of

equal value, we replicated the results from Experiment 1, as partici-

pants preferred BOGO deals (67.19%, 1075 of 1600) over percentage

deals (32.81%, 525 of 1600), which was signiﬁcantly diﬀerent from a

non-preferential 50–50 distribution, χ

= 97.41, p < .001.

4.2.3. Discount level

As in Experiment 1, to test for the magnitude of the discount, we

split the discount levels into low discounts (20% and 25%) and high

discounts (33% and 40%). We ran a generalized linear mixed model

controlling for the random eﬀect of participants, with the discount level

predicting participants' preferred deal. Replicating the results from

Experiment 1, the analysis revealed that discount level inﬂuenced

preferences, with participants being more likely to prefer BOGO deals

when the discount was high (72.63%, 581 of 800) than when it was low

(61.75%, 494 of 800), F(1, 1598) = 25.23, p < .001, b = 0.59, 95%

CI = [0.361, 0.824]. To check for robustness, we included in the model

participants' numeracy and their willingness to pay for the diﬀerent

products in the experiment. Including these measures did not change

the eﬀect of the discount level (F(1,1593) = 25.38, p < .001, b = 0.60,

95% CI = [0.364, 0.828]. Additionally, none of the control variables

signiﬁcantly aﬀected participants' preferred deal (F's < 2.49, p's >

.114).

4.2.4. Gaze behavior

Due to a technical error, eye tracking data from one participant was

not recorded. Hence, the analysis of the eye tracking data included 49

participants. We ran a generalized linear mixed model with discount

level (high vs. low), deal type (BOGO vs. percentage), and their inter-

action predicting ﬁxation count on each of the two deals in the ex-

perimental trials. The analysis revealed a main eﬀect for discount level,

F(1,3132) = 27.68, p < .001 and a main eﬀect for deal type, F

(1,3132) = 11.23, p = .001. Replicating the results of Experiment 1,

those eﬀects were qualiﬁed by deal type × discount level interaction, F

(1, 3132) = 4.89, p = .027. When the discount level was low, partici-

pants exhibited more ﬁxations on percentage deals (M = 5.31,

SD = 4.80) than on BOGO deals (M = 4.60, SD = 4.44), F(1,

3132) = 15.46, p < .001, b = 0.71, 95% CI = [0.356, 1.063].

However, when the discount level was high, participants had similar

number of ﬁxations both on percentage deals (M = 4.36, SD = 4.03)

and on BOGO deals (M = 4.21, SD = 3.83), F < 1.

As in the previous analyses, we ran an additional model including

participants' numeracy and willingness to pay for the diﬀerent products

as a robustness check. The main eﬀects of discount level (F(1,

3127) = 27.68, p < .001) and deal type (F(1, 3127) = 11.23,

p = .001), as well as the deal type × discount level interaction (F(1,

3127) = 4.89, p = .027) remained signiﬁcant, whereas none of the

control variables aﬀected the ﬁxation count (F's < 3.76, p's > .053).

5. General discussion

For companies to sell their products in a competitive market,

especially in one as the online commerce, it is crucial to develop ef-

fective ways to promote sales. Promotion deals are a powerful and easy-

to-implement tools, and nearly half of all food items are being sold

under some type of promotion deal (Bogomolova et al., 2015). Yet,

choosing how to present these deals is not trivial: For a deal to be ef-

fective, it must be implemented wisely. Our ﬁndings suggest that not all

promotion deals are equal in the amount of attention they attract.

In two experiments, we found that when a promotion deal requires

the purchase of a certain quantity of items (i.e., a multi-buy deal),

people prefer deals that are presented as giving them some of the items

for free (i.e., BOGO deals) over deals that are presented as a price re-

duction (i.e., percentage deals). Interestingly, this preference leads

some people to choose a ﬁnancially worse deal when it is presented as a

BOGO deal than when it is presented as a percentage deal. That is,

people at times seem to exhibit suboptimal preferences (Shaﬁr &

LeBoeuf, 2002) and favor a less proﬁtable deal if it is presented in a

more attractive manner. This cannot be explained by people's numeral

abilities. Numeral abilities were positively associated with accuracy in

detecting ﬁnancially better deals, but nevertheless, did not prevent

people from choosing ﬁnancially worse deals – especially when the

T. Gordon-Hecker, et al.

Journal of Business Research 111 (2020) 128–134

132

worse deal was more attractively presented, i.e., as a BOGO deal.

Previous research has shown that for larger discounts, people prefer

price reductions to package enlargement (Hardesty & Bearden, 2003).

Challenging this ﬁnding, here we found that when a percentage deal

requires a purchase of a certain quantity of items, the preference for

BOGO deals is stronger when the discount level is high (vs. low).

Consumers are less likely to process information extensively when

discount levels are high versus more moderate (Grewal, Marmorstein, &

Sharma, 1996). Hence, when purchasing a certain quantity of items is

not required, high discounts are preferable to large bonus packs, be-

cause the consumer can more easily understand and process a discount

(Diamond & Campbell, 1989; Hardesty & Bearden, 2003). However,

when consumers must purchase a certain quantity of an item to enjoy

the price discount, that percentage promotion is as complex as a BOGO

deal. Therefore, when the discount level is high, and consumers are less

invested in processing the information, they may follow their initial

preferences and choose BOGO deals over percentage deals.

Complementing the interpretation of diﬀerent information processes

as the cause for the eﬀect of discount levels on the preferred deal, the

current research also reveals that consumers' attention allocation to the

diﬀerent types of deals is inﬂuenced by the magnitude of the discount.

Generally, percentage deals are more diﬃcult to process than BOGO

deals (Tan & Bogomolova, 2016). Indeed, when discount level was low,

which as described above should lead to more thorough information

processing (Grewal et al., 1996), people exhibited more ﬁxations on

percentage deals than on BOGO deals, suggesting that percentage deals

indeed required more re-ﬁxations (an indicator of attention) to be suf-

ﬁciently processed and understood (Rayner, 2009). When the discount

level was high, however, and participants presumably processed in-

formation less thoroughly, their attention was drawn to the BOGO deals

– people no longer ﬁxated more often on percentage deals than on BOGO

deals (Experiment 2) or even ﬁxated more often on BOGO deals than

percentage deals (Experiment 1). This increase in ﬁxations on BOGO

deals when discount levels were high was associated with an increase in

choosing BOGO deals over discount deals.

One interesting question not addressed by our research is the pos-

sible moderating role of the type of product on consumers' attention.

For instance, research has shown that people prefer percentage deals for

vice products (i.e., products that grant immediate satisfaction, e.g.,

chocolate) and BOGO deals for non-vice products (i.e., products that

grant long-term payoﬀs, e.g., vegetables; Wertenbroch, 1998), arguably

in order to refrain from consuming more of a vice product (Mishra &

Mishra, 2011). Interestingly, in our second experiment participants

were exposed to only vice products (i.e., chocolate bars), yet, we found

a preference for BOGO deals over percentage deals, especially when

discount levels were high. Of course, in our experiment, both types of

deals required the purchase of the quantity of products, making the

motivation to stay away from BOGO deals in order to refrain from

consuming more of a vice product irrelevant. The question is, whether

the preference we found for BOGO deals would even be stronger for

non-vice products, or if an equal quantity of products in deals undoes

diﬀerences in preference for deal type for vice and non-vice products.

6. Conclusion

Customers often spend minimal eﬀort on processing information

regarding prices and discounts, especially when discount levels are high

(Chen, Monroe, & Lou, 1998; Grewal et al., 1996; Hardesty & Bearden,

2003). Here, we show that when discounts are high, consumers' at-

tention shifts more toward BOGO deals than toward percentage deals,

and that this is accompanied by an increase in choice for BOGO deals

over discount deals. In today's world of online commerce, when con-

sumers are exposed to hundreds of products, deals, and promotions, the

ability of retailers to attract potential consumers' attention is key. By

oﬀering the most attractive deal, companies may be able to increase

their sales and revenues. Consumers, at the other end, may realize that

it is worthwhile to more thoroughly evaluate oﬀers when they under-

stand how retailers can steer their attention by presenting tempting

BOGO deals.

Appendix A. Complete list of trials in Experiments 1 & 2 (each combination was presented 4 times in each experiment, each with

repetition with diﬀerent products)

Deal on the left Deal on the right

Buy 2 get 1 for free Buy 3 get 33% oﬀ

Buy 2 get 1 for free Buy 4 get 25% oﬀ

Buy 2 get 1 for free Buy 5 get 40% oﬀ

Buy 3 get 1 for free Buy 3 get 33% oﬀ

Buy 3 get 1 for free Buy 4 get 25% oﬀ

Buy 3 get 1 for free Buy 5 get 20% oﬀ

Buy 3 get 2 for free Buy 3 get 33% oﬀ

Buy 3 get 2 for free Buy 5 get 20% oﬀ

Buy 3 get 2 for free Buy 5 get 40% oﬀ

Buy 4 get 1 for free Buy 4 get 25% oﬀ

Buy 4 get 1 for free Buy 5 get 20% oﬀ

Buy 4 get 1 for free Buy 5 get 40% oﬀ

Buy 3 get 33% oﬀ Buy 2 get 1 for free

Buy 3 get 33% oﬀ Buy 3 get 1 for free

Buy 3 get 33% oﬀ Buy 3 get 2 for free

Buy 4 get 25% oﬀ Buy 2 get 1 for free

Buy 4 get 25% oﬀ Buy 3 get 1 for free

Buy 4 get 25% oﬀ Buy 4 get 1 for free

Buy 5 get 20% oﬀ Buy 3 get 1 for free

Buy 5 get 20% oﬀ Buy 3 get 2 for free

Buy 5 get 20% oﬀ Buy 4 get 1 for free

Buy 5 get 40% oﬀ Buy 2 get 1 for free

Buy 5 get 40% oﬀ Buy 3 get 2 for free

Buy 5 get 40% oﬀ Buy 4 get 1 for free

T. Gordon-Hecker, et al.

Journal of Business Research 111 (2020) 128–134

133

Appendix B. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.jbusres.2019.02.070.

References

Balcetis, E., & Dunning, D. (2006). See what you want to see: Motivational inﬂuences on

visual perception. Journal of Personality and Social Psychology, 91(4), 612–625.

Bee, N., Prendinger, H., Nakasone, A., André, E., & Ishizuka, M. (2006). Autoselect: What

you want is what you get: Real-time processing of visual attention and aﬀect. In E.

André, L. Dybkjaer, W. Minker, H. Neumann, & M. Weber (Eds.). Perception and in-

teractive technologies (pp. 40–52). Berlin, Germany: Springer.

Blattberg, R. C., & Neslin, S. A. (1989). Sales promotion: The long and the short of it.

Marketing Letters, 1(1), 81–97.

Bogomolova, S., Dunn, S., Trinh, G., Taylor, J., & Volpe, R. J. (2015). Price promotion

landscape in the US and UK: Depicting retail practice to inform future research

agenda. Journal of Retailing and Consumer Services, 25, 1–11.

Boland, W. A., Connell, P. M., & Erickson, L. M. (2012). Children's response to sales

promotions and their impact on purchase behavior. Journal of Consumer Psychology,

22(2), 272–279.

Chandon, P., Hutchinson, J. W., Bradlow, E. T., & Young, S. H. (2009). Does in-store

marketing work? Eﬀects of the number and position of shelf facings on brand at-

tention and evaluation at the point of purchase. Journal of Marketing, 73(6), 1–17.

Chen, S. F. S., Monroe, K. B., & Lou, Y. C. (1998). The eﬀects of framing price promotion

messages on consumers' perceptions and purchase intentions. Journal of Retailing,

74(3), 353–372.

Diamond W. D., & Campbell L. (1989), The Framing of Sales Promotions: Eﬀects on

Reference Price Change. In: Srull T.K., (ed). NA - Advances in Consumer Research

Volume 16 (p. 241–247), Provo, UT : Association for Consumer Research.

Diamond, W. D., & Sanyal, A. (1990),The Eﬀect of Framing on the Choice of Supermarket

Coupons. In Goldberg, M.E., Gorn, G., & Pollay, R. W. (Eds.), NA - Advances in

Consumer Research Volume 17 (pp. 488–493), Provo, UT : Association for Consumer

Research.

Drèze, X., & Hussherr, F. X. (2003). Internet advertising: Is anybody watching? Journal of

Interactive Marketing, 17(4), 8–23.

Elliot, A. J., & Thrash, T. M. (2010). Approach and avoidance temperament as basic di-

mensions of personality. Journal of Personality, 78(3), 865–906.

Epstein, S., Pacini, R., Denes-Raj, V., & Heier, H. (1996). Individual diﬀerences in in-

tuitive–experiential and analytical–rational thinking styles. Journal of Personality and

Social Psychology, 71(2), 390–405.

Fiedler, S., Glöckner, A., Nicklisch, A., & Dickert, S. (2013). Social value orientation and

information search in social dilemmas: An eye-tracking analysis. Organizational

Behavior and Human Decision Processes, 120(2), 272–284.

Gable, P., & Harmon-Jones, E. (2010). The blues broaden, but the nasty narrows:

Attentional consequences of negative aﬀects low and high in motivational intensity.

Psychological Science, 21(2), 211–215.

Gal, D., & Rucker, D. D. (2018). The loss of loss aversion: Will it loom larger than its gain?

Journal of Consumer Psychology, 28(3), 497–516.

Glöckner, A., Fiedler, S., Hochman, G., Ayal, S., & Hilbig, B. E. (2012). Processing dif-

ferences between descriptions and experience: A comparative analysis using eye-

tracking and physiological measures. Frontiers in Psychology, 3, 173.

Grewal, D., Krishnan, R., Baker, J., & Borin, N. (1998). The eﬀect of store name, brand

name and price discounts on consumers' evaluations and purchase intentions. Journal

of Retailing, 74(3), 331–352.

Grewal, D., Marmorstein, H., & Sharma, A. (1996). Communicating price information

through semantic cues: The moderating eﬀects of situation and discount size. Journal

of Consumer Research, 23(2), 148–155.

Hardesty, D. M., & Bearden, W. O. (2003). Consumer evaluations of diﬀerent promotion

types and price presentations: The moderating role of promotional beneﬁt level.

Journal of Retailing, 79(1), 17–25.

Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk.

Econometrica, 47(2), 263–292.

Krishna, A., Briesch, R., Lehmann, D. R., & Yuan, H. (2002). A meta-analysis of the impact

of price presentation on perceived savings. Journal of Retailing, 78(2), 101–118.

Lattin, J. M., & Bucklin, R. E. (1989). Reference eﬀects of price and promotion on brand

choice behavior. Journal of Marketing Research, 299–310.

Mazar, N., Shampanier, K., & Ariely, D. (2016). When retailing and Las Vegas meet:

Probabilistic free price promotions. Management Science, 63(1), 250–266.

Meißner, M., Musalem, A., & Huber, J. (2016). Eye tracking reveals processes that enable

conjoint choices to become increasingly eﬃcient with practice. Journal of Marketing

Research, 53(1), 1–17.

Mishra, A., & Mishra, H. (2011). The inﬂuence of price discount versus bonus pack on the

preference for virtue and vice foods. Journal of Marketing Research, 48(1), 196–206.

Mogg, K., Bradley, B. P., Field, M., & De Houwer, J. (2003). Eye movements to smoking-

related pictures in smokers: Relationship between attentional biases and implicit and

explicit measures of stimulus valence. Addiction, 98(6), 825–836.

Orquin, J. L., & Holmqvist, K. (2018). Threats to the validity of eye-movement research in

psychology. Behavior Research Methods, 50(4), 1645–1656.

Orquin, J. L., & Mueller-Loose, S. (2013). Attention and choice: A review on eye move-

ments in decision making. Acta Psychologica, 144(1), 190–206.

Papies, E. K., Stroebe, W., & Aarts, H. (2008). Healthy cognition: Processes of self-reg-

ulatory success in restrained eating. Personality and Social Psychology Bulletin, 34(9),

1290–1300.

Pieters, R., & Wedel, M. (2004). Attention capture and transfer in advertising: Brand,

pictorial, and text-size eﬀects. Journal of Marketing, 68(2), 36–50.

Pittarello, A., Leib, M., Gordon-Hecker, T., & Shalvi, S. (2015). Justiﬁcations shape ethical

blind spots. Psychological Science, 26(6), 794–804.

Raghubir, P. (2004). Free gift with purchase: Promoting or discounting the brand? Journal

of Consumer Psychology, 14(1–2), 181–186.

Rayner, K. (2009). Eye movements and attention in reading, scene perception, and visual

search. Quarterly Journal of Experimental Psychology, 62(8), 1457–1506.

Reichle, E. D., Rayner, K., & Pollatsek, A. (2003). The EZ reader model of eye-movement

control in reading: Comparisons to other models. Behavioral and Brain Sciences, 26(4),

445–476.

Shaﬁr, E., & LeBoeuf, R. A. (2002). Rationality. Annual Review of Psychology, 53(1),

491–517.

Shampanier, K., Mazar, N., & Ariely, D. (2007). Zero as a special price: The true value of

free products. Marketing Science, 26(6), 742–757.

Shimojo, S., Simion, C., Shimojo, E., & Scheier, C. (2003). Gaze bias both reﬂects and

inﬂuences preference. Nature Neuroscience, 6(12), 1317–1322.

Sinha, I., & Smith, M. F. (2000). Consumers' perceptions of promotional framing of price.

Psychology & Marketing, 17(3), 257–275.

Tan, P. J., & Bogomolova, S. (2016). A descriptive analysis of consumer's price promotion

literacy skills. International Journal of Retail & Distribution Management, 44(12),

1223–1244.

Townshend, J., & Duka, T. (2001). Attentional bias associated with alcohol cues:

Diﬀerences between heavy and occasional social drinkers. Psychopharmacology,

157(1), 67–74.

Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases.

science, 185(4157), 1124–1131.

U.S. Department of Commerce. Quarterly Retail E-Commerce Sales: 2nd Quarter 2017.

Washington, D.C 20233. August 17, 2017.

Weller, J. A., Dieckmann, N. F., Tusler, M., Mertz, C. K., Burns, W. J., & Peters, E. (2013).

Development and testing of an abbreviated numeracy scale: A Rasch analysis ap-

proach. Journal of Behavioral Decision Making, 26(2), 198–212.

Wertenbroch, K. (1998). Consumption self-control by rationing purchase quantities of

virtue and vice. Marketing Science, 17(4), 317–337.

T. Gordon-Hecker, et al.

Journal of Business Research 111 (2020) 128–134

134